Seaborn lineplot using median instead of mean

Question

I'm using seaborn.lineplot() to create a line figure like this (a line representing mean, surrounded by a band representing the std):

sns.lineplot(x="trial", y="rvalues", hue="subject", err_style="band", ci='sd', data=df)

My only issue is that since my data is not Gaussian, I care more about the median instead of mean. How to do that in Seaborn?

Or are there similar tools that are capable? I know I can do from scratch in matplotlib but that requires a lot of work to make it this nice.

Answer 1

estimator should be a pandas method.

Use estimator=np.median instead of estimator="median".

sns.lineplot(x="trial", y="rvalues", hue="subject", err_style="band",
             ci='sd',
             estimator=np.median,
             data=df)

Answer 2

From the documentation:

estimator : name of pandas method or callable or None, optional
Method for aggregating across multiple observations of the y variable at the same x level. If None, all observations will be drawn.

Hence try

sns.lineplot(x="trial", y="rvalues", hue="subject", err_style="band", 
             ci='sd', estimator="median", data=df)

Here is a minimal example:

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

df = pd.DataFrame({"trial" : np.random.randint(10, size=350),
                   "rvalues" : np.random.randn(350),
                   "subject" : np.random.randint(4, size=350)})

sns.lineplot(x="trial", y="rvalues", hue="subject", err_style="band", 
             ci='sd', estimator="median", data=df)
plt.show()